Wetlands of the Prairie Pothole Region of North America provide habitat for over 50% of the continent's breeding waterfowl, but most of the region's wetlands have been lost or degraded through intensive agricultural development. Despite widespread wetland losses in much of the Canadian prairies, there is little information about trends in degradation of remaining wetlands. Using habitat data collected for ~10,500 wetlands across the Canadian prairies during annual waterfowl surveys, 1985-2005, I employed multistate models in Program MARK to estimate rates of impact and recovery of wetlands resulting from agricultural activities. Then, I characterized the incidence of agricultural degradation to these wetlands. Rates of impact to wetland margins (natural vegetation around flooded basins) declined over time, likely due to a decreasing percentage of unaffected wetlands; recovery rates for margins were always lower than impact rates, suggesting increased cumulative degradation of wetlands over time. Unlike margins, impact and recovery rates for basins fluctuated with spring pond densities. Shallow ephemeral wetlands located in agricultural fields had the highest impact and lowest recovery rates. Multistate modeling could also be used to estimate rates associated with other landscape processes.<p><p> My second objective was to determine whether physical characteristics of prairie Canada wetlands could be used to predict breeding duck abundance. First, I sought to determine how pre-existing models developed in the Dakotas (USA) performed when predicting breeding duck abundances on Canadian prairie wetlands. I related duck pair abundance to pond area, and then compared observed to predicted duck abundance. The Dakota models performed reasonably well in predicting numbers of blue-winged teal (<i>Anas discors</i>), gadwall (<i>A. strepera</i>), and northern pintail (<i>A. acuta</i>), but predicted fewer mallards (<i>A. platyrhynchos</i>) and northern shovelers (<i>A. clypeata</i>) than were observed on wetlands. Pond area was an important predictor of duck abundance in all models, but results were less biased and more consistent in models developed specifically for Canadian wetlands. Spatiotemporal variation in the relationship of breeding duck abundance and wetland characteristics was also affected by regional duck and pond densities. Overall, the new applications and models developed and validated in this study will be useful for wetland and waterfowl management in the Canadian prairies.

Wetlands of the Prairie Pothole Region of North America provide habitat for over 50% of the continent's breeding waterfowl, but most of the region's wetlands have been lost or degraded through intensive agricultural development. Despite widespread wetland losses in much of the Canadian prairies, there is little information about trends in degradation of remaining wetlands. Using habitat data collected for ~10,500 wetlands across the Canadian prairies during annual waterfowl surveys, 1985-2005, I employed multistate models in Program MARK to estimate rates of impact and recovery of wetlands resulting from agricultural activities. Then, I characterized the incidence of agricultural degradation to these wetlands. Rates of impact to wetland margins (natural vegetation around flooded basins) declined over time, likely due to a decreasing percentage of unaffected wetlands; recovery rates for margins were always lower than impact rates, suggesting increased cumulative degradation of wetlands over time. Unlike margins, impact and recovery rates for basins fluctuated with spring pond densities. Shallow ephemeral wetlands located in agricultural fields had the highest impact and lowest recovery rates. Multistate modeling could also be used to estimate rates associated with other landscape processes.<p><p> My second objective was to determine whether physical characteristics of prairie Canada wetlands could be used to predict breeding duck abundance. First, I sought to determine how pre-existing models developed in the Dakotas (USA) performed when predicting breeding duck abundances on Canadian prairie wetlands. I related duck pair abundance to pond area, and then compared observed to predicted duck abundance. The Dakota models performed reasonably well in predicting numbers of blue-winged teal (<i>Anas discors</i>), gadwall (<i>A. strepera</i>), and northern pintail (<i>A. acuta</i>), but predicted fewer mallards (<i>A. platyrhynchos</i>) and northern shovelers (<i>A. clypeata</i>) than were observed on wetlands. Pond area was an important predictor of duck abundance in all models, but results were less biased and more consistent in models developed specifically for Canadian wetlands. Spatiotemporal variation in the relationship of breeding duck abundance and wetland characteristics was also affected by regional duck and pond densities. Overall, the new applications and models developed and validated in this study will be useful for wetland and waterfowl management in the Canadian prairies.

<p> The Peter Strides Pond study area, approximately 150 km^2, is located in the Southern Long Range Mountains in southwestern Newfoundland. It lies at the southern margin of the Paleozoic Central Mobile Belt in the Dunnage tectonostratigraphic zone. A felsic intrusion, granoblastic gneiss, heterolithic conglomerate and basalt form foliated, linear bodies across the study area parallel to the regional, northeast striking, structural grain; gabbro, diabase and rhyolite have non-linear exposure and no internal fabric. Two parallel mylonite zones traverse the study area. The Victoria River Shear Zone (VRSZ) to the north and the Peter Strides Pond Shear Zone (PSPSZ) to the south are concordant with the regional fabric and separate domains which increase in metamorphic grade from lower greenschist in the northwest to upper amphibolite facies to the southeast. Veins and lenses of variable form and composition are observed in several lithologies but the focussed occurrence of milky white, foliation parallel quartz veins in both the VRSZ and the felsic pluton is significant to these units.</p> <p> A tentative stratigraphic succession for the study area utilizes regional correlations, radiometric ages and fossil evidence because limited exposure does not reveal contacts between adjacent lithologies. Basalt of the Victoria Lake Group and gneiss of the Bay du Nord Group are the oldest units in the area. Conglomerate unconformably overlies the Victoria Lake Group and contains volcanic and sedimentary clasts derived from the underlying group. Mafic to felsic plutonic igneous rocks intrude volcanic rocks of the Victoria Lake Group. Undated gabbro, diorite and diabase may be coeval with Devonian adamellite-granodiorite. Parallel shear zones are concordant with the regional structural grain and follow lithologic contacts closely. Megacrystic granite cross-cuts PSP mylonite, adamellite and Bay du Nord gneiss. It is the youngest unit in the study area. The study area's tectonic framework represents a compressional environment which dominated during, and continued after Iapetus closure. Mafic volcanic rocks adjacent to ophiolite complexes represent back arc basins preserved via obduction. Continued compression was accomodated by crustal thickening through reverse, northwest directed, thrust faults. The inhomogeneously thickened crust provided a host of potential magma compositions to be intruded locally.</p> / Thesis / Master of Science (MSc)